Bigfoot: True or Real?

An explainer on how to use simple functions from ggplot and leaflet.

---

Intro

I recently re-watched the iCarly episode “iBelieve in Bigfoot,” where the cast tries to prove the legendary apelike figure, claimed to have been seen by thousands, is real.

During the episode, A bigfoot-expert, Dr. Sydney Van Gurbin, mentions he is going hunting to find the creature in Mount Baker National Forest. According to the iCarly show fandom page, Mount Baker National Forest is a forest in the episode where multiple sightings of Bigfoot took place.

In real life, Mount Baker was a national forest in Washington that existed between 1924 and 1974. It was later combined with Snoqualmie National Forest to form Mount Baker-Snoqualmie National Forest.

This led me to begin pondering at the thought of the actual existence of Bigfoot. Is it just some random person in a suit walking around to spook people or is it actually a mythical creature lurking in the woods?

It sounds silly, yet, at the same time, I was excited to learn where people encountered him the most.

Dr. Sydney Van Gurbin appeared for the first time on iCarly to talk about his Bigfoot sightings. May 8, 2010.

I began by stumbling upon a dataset of Bigfoot sightings by a data scientist named Timothy Renner. He originally got the data from The Bigfoot Field Researchers Organization who have compiled a huge list of verified Bigfoot sightings throughout the United States and North America. The organization made the data publicly available and Renner kindly cleaned it up for use.

bigfoot_sightings %>%
  glimpse()

## Rows: 5,021
## Columns: 28
## $ observed           <chr> "I was canoeing on the Sipsey river in Alabama. It ~
## $ location_details   <chr> "", "East side of Prince William Sound", "Great swa~
## $ county             <chr> "Winston County", "Valdez-Chitina-Whittier County",~
## $ state              <chr> "Alabama", "Alaska", "Rhode Island", "Pennsylvania"~
## $ season             <chr> "Summer", "Fall", "Fall", "Summer", "Spring", "Fall~
## $ title              <chr> "", "", "Report 6496: Bicycling student has night e~
## $ latitude           <dbl> NA, NA, 41.45000, NA, NA, 35.30110, 39.38745, 41.29~
## $ longitude          <dbl> NA, NA, -71.50000, NA, NA, -99.17020, -81.67339, -7~
## $ date               <chr> "", "", "1974-09-20", "", "", "1973-09-28", "1971-0~
## $ number             <dbl> 30680, 1261, 6496, 8000, 703, 9765, 4983, 31940, 56~
## $ classification     <chr> "Class B", "Class A", "Class A", "Class B", "Class ~
## $ geohash            <chr> "", "", "drm5ucxrc0", "", "", "9y32z667yc", "dpjbj6~
## $ temperature_high   <dbl> NA, NA, 78.17, NA, NA, 71.86, NA, 92.24, NA, NA, 74~
## $ temperature_mid    <dbl> NA, NA, 73.425, NA, NA, 61.425, NA, 80.810, NA, NA,~
## $ temperature_low    <dbl> NA, NA, 68.68, NA, NA, 50.99, NA, 69.38, NA, NA, 53~
## $ dew_point          <dbl> NA, NA, 65.72, NA, NA, 51.03, NA, 67.34, 32.55, NA,~
## $ humidity           <dbl> NA, NA, 0.86, NA, NA, 0.79, NA, 0.68, 0.45, NA, 0.7~
## $ cloud_cover        <dbl> NA, NA, 0.86, NA, NA, 0.11, NA, 0.05, 0.00, NA, 0.6~
## $ moon_phase         <dbl> NA, NA, 0.16, NA, NA, 0.07, NA, 0.76, 0.02, NA, 0.1~
## $ precip_intensity   <dbl> NA, NA, 0.0000, NA, NA, NA, NA, 0.0000, 0.0000, NA,~
## $ precip_probability <dbl> NA, NA, 0.00, NA, NA, NA, NA, 0.00, 0.00, NA, 0.70,~
## $ precip_type        <chr> "", "", "", "", "", "rain", "", "", "", "", "rain",~
## $ pressure           <dbl> NA, NA, 1020.61, NA, NA, 1017.26, NA, 1016.80, 1012~
## $ summary            <chr> "", "", "Foggy until afternoon.", "", "", "Partly c~
## $ uv_index           <int> NA, NA, 4, NA, NA, 7, NA, 8, 8, NA, 6, 10, 6, 7, NA~
## $ visibility         <dbl> NA, NA, 2.750, NA, NA, 10.000, NA, 6.922, 8.880, NA~
## $ wind_bearing       <int> NA, NA, 198, NA, NA, 259, NA, 219, 285, NA, 262, 19~
## $ wind_speed         <dbl> NA, NA, 6.92, NA, NA, 8.41, NA, 1.01, 4.01, NA, 0.4~

Where are you Bigfoot?

To figure out where Bigfoot could potentially be I knew I had to narrow down his “sighting” to a few states. I decided to summarise the total amount of sightings by state from 1940 to 2021. From there, I was able to confirm the top 10 states that reported sightings the most.

Those states were Washington, California, Ohio, Florida, Oregon, Illinois, Texas, Michigan, Missouri, and Colorado.

# Top five states with the most Bigfoot sightings from 1940 to 2021.

top10 <-
bigfoot_sightings %>%

  group_by (state) %>%
  summarise (times_spotted = n()) %>%

  arrange (desc(times_spotted)) %>%
  head (10)

top10

Honestly, I was excited to see Washington as number one on the list. It felt like there was some truth to the iCarly episode I watched.

Next, I wanted to show this data I found on a visually appealing scatter plot using the package ggplot2.

ggplot2 is a tidyverse data visualization package that is a system for creating graphics, based on Leland Wilkinson’s 1999 book “Grammar of Graphics.” You provide the data, tell ggplot2 how to map variables to aesthetics, what graphical primitives to use, and it takes care of the details.

---

Here is how I used it:

create new variable <- ggplot ( data = what data do I want to look at, aes (x= column in my data I can compare to y , y = column in my data I can compare to x)) +

geom_point() + theme_minimal() + aes (size= what do I want to measure by) + aes (color = what column(s) do I want to have color) + geom_point (alpha = 0.4) + ggtitle (‘title of plot map’, subtitle = “subtitle of plotmap.”)

ggplotly(new variable I created at top) This makes the plot map interactive after knitting.

Important terms and notes:

• aesthetics (called “aes”) : What data are you plotting? It can have more than two dimensions, such as x-axis, y-axis, a column used for color and another used for size. These are things we can see on a graphic, including the values. -geometry refers to the shape each column will take.
• scale - any transformation we might make.
• facets - creating small multiples by splitting up one graph into many similar graphs, based on a variable.
• layers - adding multiple geometries on top of one another to reveal new information, or add annotations.
• instead of the %>% pipe command, it uses + instead.
• geom_point is used to create scatterplots.

---

us_bf <-
  ggplot (
    data = top10,
    aes (x= times_spotted , y = state)
  ) +  geom_point() +theme_minimal() + aes (size=times_spotted) + aes (color = state) +
  geom_point (alpha = 0.4) + ggtitle('Bigfoot Sightings in the United States', subtitle = "Total amount of times BF has been sighted (1940 to 2021).")

ggplotly(us_bf)

Next, I want to create a similar scatterplot but only with information regarding sightings from counties in Washington State. However, to do this, I went ahead and filtered/summarised data for the top 10 counties in Washington that reported the most Bigfoot sightings from 1940 to 2021.

#Note: some counties are missing information in the year column. 

# Here We filtered to only look at the state of Washington. 
bigfoot_wa <-
bigfoot_sightings %>%
  filter (state == "Washington") %>%
  mutate (year = str_sub (date ,1,4)) %>%

  
  # I am using the following "select" code because the one we usually use does not not apply in this R markdown. It keeps trying to read "select" from the raster package instead of the tidyverse package. 
  dplyr::select(year, state, county, observed, longitude, latitude)


  
  
 
bf_spotted <-  
bigfoot_wa %>%
  group_by (year, county, state) %>%
  summarise (times_spotted = n ())


bf_spotted

wa_bf_county <-
bf_spotted %>%
 group_by (county) %>%
  summarise (times_spotted = n()) %>%

  arrange (desc(times_spotted)) %>%
  head (10)

wa_bf_county

I was able to confirm the top 10 counties that reported sightings the most.

Those counties were Pierce County, Snohomish County, King County, Skamania County, Lewis County, Grays Harbor County, Jefferson County, Mason County, Okanogan Countym Chelan County.

county_bf <-
  ggplot (
    data = wa_bf_county,
    aes (x= times_spotted , y = county)) + 
    geom_point() + 
    theme_minimal() + 
   aes (size=times_spotted) +
   aes (color = county) +
   geom_point (alpha = 0.4) +
   ggtitle('Bigfoot Sightings in the United States', subtitle = "Total amount of times BF has been sighted (1940 to 2021).")

ggplotly(county_bf)

‘made me pause with fear’

Given the fact that the perimeter surrounding Mount Baker and Whatcom County being declared a “Sasquatch protection and refuge area,” I am a little surprised it did not make the top ten list.

Nevertheless, I used the package Leaflet to create an overall map of where in Washington Bigfoot was sighted from 1940 to 2021.

Leaflet is a free and open-source JavaScript library and Application Programming Interface (API) for developing interactive web maps

---

This is how I used it:

leaflet() %>% addProviderTiles(“type of base map I want to see here”) %>% addMarkers(data = data I want to pull from, popup = ~ what column I want to see when I hover over marks) %>%

setView(lat= the latitude of map , lng= the longitude of map, zoom= how close you want to see details of a specific lat and lng.)

Important terms and notes:

• There many base maps available. You can find them here and here.

---

leaflet() %>%
addProviderTiles("Esri.NatGeoWorldMap") %>%
  addMarkers(data = bigfoot_wa, popup = ~ observed) %>%
  
setView(lat= 47.5, lng=-120.7, zoom=7)

I thought it was very cool because Leaflet gave me the option to be able to included the experiences that came with each data point.

You got to actual hear from people who potentially encountered the state’s official ‘cryptid,’ whether it was a good experience or a bad one.

I leave it here for now and hopefully one day I can figure out the truth of Bigfoot.

Data

This explainer uses data from the The Bigfoot Field Researchers Organization, Timothy Renner, and Dark Sky API.

The Bigfoot Field Researchers Organization (BFRO) Founded in 1995, the BFRO is now the oldest and largest organization of its kind – a virtual community of scientists, journalists, and specialists from diverse backgrounds. The researchers who compose the BFRO are engaged in projects, including field and laboratory investigations, designed to address various aspects of the bigfoot phenomenon. As a result of the education and experience of its members and the quality of their efforts, the BFRO is widely considered as the most credible and respected investigative network involved in the study of this subject.

The data can be found here: http://www.bfro.net/GDB/

Timothy Renner Renner is a data scientist who cleaned the data found on BFRO.

The data can be found and downloaded here: https://data.world/timothyrenner/bfro-sightings-data